Third, Allan; Bennett, Brandon and Mallenby, David
Automatic grounding of vague geographic ontology in data.
In: Geocomputation 2007, 03-05 September 2007, Maynooth, Ireland.
Full text available as:
In constructing an ontological theory of a domain such as geography, it is important not only to take account of the vagueness and ambiguity which is inherent in many of the relevant concepts, but also to be able to relate the high-level definitions of the theory to actual sets of data of varying kinds. Any attempt to ignore or remove vagueness and ambiguity risks errors and conflict in the ontological theory with the knowledge of different domain experts, while an inability to ground the theory in real data limits its practical use. We present here a means of structuring such a theory to handle these issues in a principled manner, which lends itself to concrete implementation. We illustrate with reference to several examples from the domain of hydrography.
Actions (login may be required)